September 2, 2020 Ms. Marlene H. Dortch Secretary Federal Communications Commission Office of the Secretary 445 12th Street, SW Room TW-A325 Washington, DC 20554 Re: Comment on NTIA Petition for Rulemaking, Docket No. RM-11862 I. Statement of Interest NetChoice is dedicated to the promotion of consumer choice, competition, and free enterprise on the internet. Our membership includes a wide variety of e-commerce models, and businesses of all sizes. We submit these comments with these priorities in mind, and on the basis of our two decades of experience advocating for freedom of expression and enterprise online. As is our custom, we add that the views expressed herein are those of NetChoice and do not necessarily represent the views of all NetChoice members.

II. Executive Summary

The Federal Communications Commission (“FCC” or “Commission”) should dismiss the petition for rulemaking (the “Petition”) filed by the National Telecommunications and Information Administration (“NTIA”) regarding Section 230 of the Communications Decency Act (“Section 230”).1 The Petition should be dismissed for the following reasons:

(1) Congress did not intend Section 230 to create FCC regulatory authority over the internet. Both the language of the statute and its legislative history make this clear.

(2) After two decades of judicial interpretation of Section 230, during which the FCC has affirmatively chosen not to assert jurisdiction to regulate online speech, the FCC should not now seek through rule to re-interpret the statute in novel ways. Having properly not done so in the 24 years since the statute was enacted, the Commission has no special authority or expertise to which courts would defer.

1 47 U.S.C. § 230.

2

(3) The proposed re-interpretations of Section 230 that the Petition is urging the FCC to adopt via rulemaking would violate the plain language of the statute.

(4) As made clear by the Commission’s long-standing policy and practice of abjuring interpretive jurisdiction over Section 230 and its subject matter, it is questionable that an assertion of such jurisdiction at this late date would be upheld by the courts. Without need of resolving that question, it is unquestionably within the FCC’s power to exercise its discretion to reject the Petition. The Commission should do so.

III. Full Comments of NetChoice

A. Congress Did Not Intend Section 230 to Create FCC Regulatory Authority Over the Internet

The Petition asserts that the FCC has the authority to promulgate rules re-interpreting Section 230 differently than the courts have interpreted the statute over the last 24 years since it was enacted. The Petition advances three reasons for this assertion. Each is faulty.

1. Statutory Text and Legislative History First, the Petition claims:

Neither section 230’s text, nor any speck of legislative history, suggests any congressional intent to preclude the Commission’s implementation.2

This is false. In fact, the very opposite is true. Congress was emphatic that it was not creating new regulatory authority for the FCC when it enacted Section 230. This is as clear from the face of the statute as it is from the legislative history. One of the signatories to this comment, former Representative Christopher Cox (R-CA), is the author, and original co-sponsor with then-Representative Ron Wyden (D-OR), of Section 230.3 When this legislation came to the floor of the House of Representatives for debate on August 4, 1995, Representatives Cox and Wyden, together with members on both sides of the aisle, explained that their purpose was to ensure that the FCC would not have regulatory authority over content on the internet. They decried the unwelcome pro-regulatory alternative of giving the FCC responsibility for regulating content on the

2 Petition at 17. 3 For a detailed discussion of the legislative history of Section 230, see Christopher Cox, The Origins and Original Intent of Section 230 of the Communications Decency Act, Rich. J. L. & Tech. Blog (2020), https://jolt.richmond.edu/2020/08/27/the-origins-and-original-intent-of-section-230-of-the-communications-decency-act/ (last visited Aug 27, 2020).

3

internet, which at the time was being advanced in separate legislation by Senator James Exon (D-NE).

The Cox-Wyden bill under consideration was intended as a rebuke to that entire concept. Then-Representative Christopher Cox put the matter succinctly:

Some have suggested, Mr. Chairman, that we take the Federal Communications Commission and turn it into the ‘Federal Computer Commission’ — that we hire even more bureaucrats and more regulators who will attempt, either civilly or criminally, to punish people by catching them in the act of putting something into cyberspace. Frankly, there is just too much going on on the Internet for that to be effective….

The message today should be, from this Congress: we embrace this new technology, we welcome the opportunity for education and political discourse that it offers for all of us. We want to help it along this time by saying Government is going to get out of the way and let parents and individuals control it rather than Government doing that job for us.

And further from Representative Cox:

This bill will establish as the policy of the United States that we do not wish to have content regulation by the Federal Government of what is on the Internet —that we do not wish to have a ‘Federal Computer Commission’ with an army of bureaucrats regulating the Internet.

The text of Section 230 makes this explicit. Section 230(b) provides:

It is the policy of the United States … to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation.4

The bill’s author concluded:

If we regulate the Internet at the FCC, that will freeze or at least slow down technology. It will threaten the future of the Internet. That is why it is so important that we not have a ‘Federal Computer Commission’ do that.5

4 Emphasis added. 5 104th Cong., 1st Sess., 141 Cong. Rec. Part 16, 22044-45, 22047 (August 4, 1995) (remarks of Representative Cox). The FCC has acknowledged this aspect of Section 230’s legislative history. “The congressional record reflects that the drafters of section 230 did ‘not wish to have a Federal Computer Commission with an army of bureaucrats regulating the Internet.’” Federal Communications Commission, In the Matter of Restoring Internet Freedom

4

Speaker after speaker who rose in support of the Cox-Wyden measure not only extolled the bill before them, but also condemned the FCC regulatory approach being urged by Senator Exon. Representative Zoe Lofgren (D-CA) was blunt: “Senator Exon’s approach is not the right way … it will not work.” It was, she said, “a misunderstanding of the technology.”6 In the end, not a single Representative spoke against the bill or in support of a regulatory role for the FCC. The final roll call on the Cox-Wyden legislation was 420 yeas, 4 nays.7 Far from the Petition’s contention that not “a speck of legislative history” exists to show that Congress intended to keep the FCC out of this area, both the legislative history and the clear language of the law itself make it abundantly clear that this is precisely what Congress intended. And what it enacted. Congress not only did not give the Commission authority to regulate the internet in Section 230, but it expressly intended this law to prevent that result.

2. The Section 230 Definition of ‘Interactive Computer Service’ Second, the Petition falsely states that:

Section 230(f)(2) explicitly classifies “interactive computer services” as “information services.”8

To see that this is not so, one need only read the statute. Section 230(f)(2) defines “interactive computer service.”9 This was a brand-new term in 1995, when Representative Cox wrote the bill that later became Section 230 (H.R. 1978, the Internet Freedom and Family Empowerment Act). The purpose of creating a novel, bespoke term was to avoid creating jurisdiction for the FCC, in keeping with the legislative intent to keep the FCC out of this area and ensuring that “FCC” would not come to stand for “Federal Computer Commission.”

(released Jan. 4, 2018), WC Docket No. 17-108, 83 FR 7852 (02/22/2018), note 235; available at https://docs.fcc.gov/public/attachments/FCC-17-166A1.pdf (last visited Aug. 31, 2020). 6 141 Cong. Rec. Part 16, supra, at 22046. 7 Id. at 22054. 8 Petition at 47. 9 47 U.S.C. § 230(f)(2). It provides: “The term ‘interactive computer service’ means any information service, system, or access software provider that provides or enables computer access by multiple users to a computer server, including specifically a service or system that provides access to the Internet and such systems operated or services offered by libraries or educational institutions.”

5

Section 230(f)(2) created a functional definition. It embraces any service accessed by multiple users via its server. (And so, early on, the courts interpreted the definition to include almost any garden-variety website.)10 While the definition expressly includes an FCC-regulated “information service” if it otherwise answers to the description in Section 230(f)(2), it also includes millions of websites that in no way can be deemed FCC-regulated “information services.” Furthermore, “information services” are provided “via telecommunications,”11 and the Commission has determined that broadband internet access services are not telecommunications services.12 If it were true, as the Petition has asserted, that Section 230(f)(2) explicitly classifies “interactive computer services” as “information services,” then the NTIA’s jurisdictional claim would have merit. But the claim is simply not true — as is plainly evident on the face of the statute. Information services for purposes of the Communications Act of 1934 are a distinct concept from “interactive computer services” as defined in Section 230.

3. Section 201(b) Common Carrier Regulatory Authority Third, the Petition argues that:

Section 201(b) of the Communications Act (Act) empowers the Commission to … promulgate rules to resolve ambiguities in Section 230.13

Section 201 of the Communications Act sets out the FCC’s authority to regulate common carriers. This is the reference point for interpreting the final sentence of Section 201(b). Properly read in context, Section 201(b) is not a source of authority for the FCC to promulgate rules of statutory interpretation governing “interactive computer services” under Section 230. The edge services within the ambit of Section 230’s definition of “interactive computer services” are mostly content providers, as opposed to the paradigm of the common carrier, paid to transport goods or data. Among the world’s 1.7 billion websites of all sizes and varieties14 —

10 See, e.g., Batzel v. Smith, 333 F.3d 1018, 1033 (9th Cir. 2003) (one-man website is “interactive computer service” as defined in Section 230(f)(2)). 11 47 U.S.C. § 153(24). 12 In the Matter of Restoring Internet Freedom, Declaratory Ruling, Report and Order, and Order, 33 FCC Rcd 311 (2018). 13 Petition at 15. 14 Source: Internet Live Stats. Available at https://www.internetlivestats.com/total-number-of-websites/ (last visited Aug. 28, 2020).

6

virtually all of which are covered by Section 230 — some 200 million are active.15 Almost none of these online content providers would fit the traditional concept of a common carrier under even the most capacious re-imagination of the FCC’s authorities. The FCC has not classified websites and other edge services as common carriers of any sort, expressly stating so in its Restoring Internet Freedom Order of January 4, 2018: “[W]e need not and do not address … the specific category or categories into which particular edge services fall.”16 Similarly, the Commission’s 2010 Report and Order, Preserving the Open Internet, Broadband Industry Practices,17 took pains to clarify that it applied “only to the provision of broadband Internet access service and not to edge provider activities, such as the provision of content.”18 Because the edge providers covered by Section 230 are not classified as common carriers, Section 201(b) of the Communications Act of 1934 does not empower the Commission to promulgate rules under Section 230 governing them.19

B. The FCC Possesses No Special Authority or Expertise to Alter Two Decades of Judicial Interpretation of Section 230

Whether the Commission should attempt to assert its authority to interpret Section 230, enacted a quarter century ago, is a question of both authority and discretion. As discussed above, Section 230’s text plainly does not delegate rulemaking authority to the FCC. To the contrary, it states the intent of Congress to keep the FCC, and federal regulatory agencies in general, out of this area. Section 230(b) states the law’s purpose that the internet be “unfettered by Federal or State regulation.”20

15 Id. 16 In the Matter of Restoring Internet Freedom, Declaratory Ruling, Report and Order, 33 FCC Rcd 311, FCC 17-166, WC Docket No. 17-108, released Jan. 4, 2018. 83 FR 7852 (02/22/2018) n.849; available at https://apps.fcc.gov/edocs_public/attachmatch/FCC-17-166A1.pdf (last visited Sept. 2, 2020). 17 Federal Communications Commission, In the Matter of Preserving the Open Internet, Broadband Industry Practices, GN Docket No. 09-191; WC Docket No. 07-52, released Dec. 23, 2010. 76 FR 60754 (09/30/2011), 76 FR 59192 (09/23/2011); available at https://apps.fcc.gov/edocs_public/attachmatch/FCC-10-201A1_Rcd.pdf (last visited Aug. 27, 2020). 18 Id. at ¶ 50. 19 For the same reason, the Petition’s recommendation that the FCC rely upon its authority over information services to impose transparency requirements on social media and other websites that are not classified as information services (see Petition at 57) is unsupported in law. 20 47 U.S.C. § 230(b)(2).

7

As the Ninth Circuit has noted, it was an “unusual step” for Congress to set forth its policy goals explicitly in the statute.21 The court’s inference that the authors of the law did so “to guide the interpretation of [Section 230’s] broad language” is entirely correct.22 Given this background, a Commission claim to jurisdiction based upon an implied delegation would likely face considerable skepticism in the courts. In Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., the Supreme Court famously said that when a statute “is silent or ambiguous with respect to the specific” issue at hand, courts should defer to an agency’s “reasonable” interpretation of that statute.23 A key element of Chevron and its progeny since is that this deference is extended only once it is established that a statute explicitly grants the agency rulemaking authority, while leaving other specific questions of statutory interpretation unresolved. For the FCC to issue rules under Section 230, the agency must have a “textual commitment of authority” from Congress to do so.24 The Communications Act of 1934 is not such a textual commitment. Its authorization for the FCC “to prescribe such rules and regulations as may be necessary in the public interest to carry out this chapter”25 extends only to congressionally “delegat[ed] authority to the agency to elucidate a specific provision of the statute by regulation.”26

Congress cannot be said to have intended the FCC to have power to interpret a specific provision of Section 230 if it did not delegate to the FCC any rulemaking power over Section 230.27 The FCC’s general rulemaking authority in 47 U.S.C. 154(i) is not a sufficient substitute for this; it requires that a statute grant the agency jurisdiction in the first place. As the District of Columbia Circuit has explained, section 154(i) acts as a “necessary and proper clause” — it can be exercised only if some other statute grants the FCC power.28 The general authorization in Section 201(b) of the Communications Act of 1934, 47 U.S.C. § 201(b), is subject to the same limitation.

21 Enigma Software Group U.S.A v. Malwarebytes, Inc., 946 F.3d 1040 (2019). 22 Id. 23 467 U.S. 837, 843 (1984). 24 Whitman v. American Trucking Ass’ns, Inc., 531 U.S. 457, 473 (2001). 25 47 U.S.C. §201(b). 26 Chevron, 467 U.S. at 843-44 (emphasis added). 27 This requirement is sometimes referred to as Chevron “Step Zero.” See Cass R. Sunstein, Chevron Step Zero, 92 Virginia Law Review 187 (2006). 28 Motion Picture Association of America, Inc. v. FCC, 309 F.3d 796, 806 (D.C. Cir. 2002) (quoting with approval FCC Chairman Powell’s statement to this effect).

8

Because Section 230 makes clear its intention that the internet remain free of federal or state regulation — an intention aimed at the FCC specifically, as amply demonstrated during congressional debate on the legislation — any ambiguity about FCC authority would likely be resolved by the courts in the negative. In Chevron, the Supreme Court held that when “the intent of Congress is clear, that is the end of the matter … [T]he court, as well as the agency, must give effect to the unambiguously expressed intent of Congress.”29 The judicial branch will be inclined to take the legislative branch at its word, because Congress “does not alter the fundamental details of a regulatory scheme in vague terms or ancillary provisions — it does not, one might say, hide elephants in mouseholes.”30 As the Supreme Court has noted, “Congress knows to speak in plain terms when it wishes to circumscribe, and in capacious terms when it wishes to enlarge, agency discretion.”31 In Section 230, Congress spoke in plain terms: no regulation.

As the Supreme Court stated in King v. Burwell (holding that Congress did not implicitly delegate rulemaking authority to the IRS in the Affordable Care Act, even while deciding that the individual mandate was a tax): “had Congress wished to assign that question to an agency, it surely would have done so expressly.”32 The Court in that case found it “especially unlikely” that Congress would have delegated rulemaking authority over the tax to the IRS because the agency has “no expertise in crafting health insurance policy.”33 Agency expertise is the sine qua non of judicial deference to agency decisions. In explaining why courts should defer to an administrative agency’s reasonable interpretations of the statutes it has been authorized to administer, the Supreme Court has stressed the importance of the agency’s expertise, its “full understanding of the force of the statutory policy in the given situation … [and its] more than ordinary knowledge respecting the matters subjected to agency regulations.”34 This hardly describes the FCC’s relationship to the policies set out by Section 230.

29 Chevron, 467 U.S. at 842-43. This is the so-called Chevron “Step One.” 30 Whitman, 531 U.S. at 468. 31 City of Arlington v. FCC., 569 U.S. 290, 296 (2013). 32 King v. Burwell, 576 U.S. 988, 135 S. Ct. 2480, 2488–89 (2015). 33 Id. 34 Chevron, 467 U.S. at 844.

9

The FCC’s own lack of expertise in the Section 230 realm is the direct result of its conscious decision to abjure any claim of new regulatory authority deriving from the law. The Commission’s well-founded decision to stay on the sidelines for the entire 24 years of the statute’s existence will undoubtedly undermine any argument that its expertise merits deference. Throughout that long period, the Commission has repeatedly backed away from classifying these platforms as “information services” subject to FCC regulation. As recently as 2018, the Commission stated it has no authority to regulate “interactive computer services,” the term that defines the edge providers covered by Section 230.35 It specifically identified its lack of Section 230 authority as a basis for its action:

We are not persuaded that section 230 of the Communications Act grants the Commission authority that could provide the basis for conduct rules here.36

Having continuously disclaimed ownership of either regulatory authority or responsibility at all times since enactment of the statute, it would be audacious indeed for the Commission now to claim that, on the basis of its supposed special expertise, the courts should grant deference to its interpretation of Section 230. The FCC simply cannot plausibly claim that it possesses any special expertise or “more than ordinary knowledge” of the subject matter. Courts will be skeptical, therefore, not only of the FCC’s belated assertion of its own jurisdiction over Section 230 — an assertion dependent on the dubious claim that Congress impliedly intended Section 230 to grant the FCC such authority — but even more so, of any FCC interpretation of the statute that is at odds with judicial interpretations that have developed over the course of a quarter century of case law. And yet the re-interpretation of Section 230 urged by the Petition would require the FCC to overturn long-standing judicial interpretations of the statute. (See section C., infra.) This will make it doubly difficult for the Commission to convince the courts that its rules are entitled to deference, because a sudden reversal of policy is one of the most common reasons for rejection of agency rules. Instead of meriting deference, “an agency changing its course must supply a reasoned analysis,” or its rulemaking will be invalidated as arbitrary and capricious.37 This is especially likely when the courts have already made their own determination of the most reasonable interpretation of a statute, judging that interpretation to be supported by both the statutory text and congressional intent. A “regulation of this sort is itself unlawful and receives no Chevron deference.”38

35 In the Matter of Restoring Internet Freedom (2018), supra ¶ 284. 36 Id. 37 Motor Vehicle Manufacturer’s Assoc. v. State Farm Mutual Automobile Ins. Co., 463 U.S. 29, 57 (1983). 38 Encino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2126 (2016).

10

C. The Petition’s Proposed Re-Interpretation of Section 230

Violates the Plain Language of the Statute The Petition urges the Commission to promulgate rules that would overturn long-standing judicial interpretation of Section 230. Beyond the obvious problems this would create for the millions of websites that have relied on the case law as it has developed over the last quarter century, the specific revisions to the meaning of Section 230 that the Petition proposes are at variance with the statutory text.

1. The Interaction Between Subparagraphs (c)(1) and (c)(2)

The Petition proposes that the FCC promulgate a rule stating that: Section 230(c)(1) has no application to any ... action to restrict access to or availability of material provided by another information content provider or to bar any information content provider from using an interactive computer service.39

In 1995, a New York court ruled that Prodigy, one of the leading consumer internet portals of the day, could be held liable for an allegedly defamatory posting by one of its users.40 The reason the court advanced for its ruling was that Prodigy had adopted content guidelines. These requested that users refrain from posts that are “insulting” or that “harass other members” or “are deemed to be in bad taste or grossly repugnant to community standards.”41 Section 230 was Congress’s response to that decision. Section 230(c)(1) overruled the New York case by stipulating that “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” No longer could judges decide that an internet platform would be liable for its users’ content, as if the platform (and not the actual content creator) were the publisher or speaker, solely because the platform sought to enforce content guidelines.

39 Petition at 31. 40 Stratton Oakmont v. Prodigy Servs. Co., 1995 N.Y. Misc. LEXIS 229, 1995 WL 323710, 23 Media L. Rep. 1794 (N.Y. Sup. Ct. May 24, 1995). 41 Id.

11

The Petition would have the FCC rewrite Section 230(c)(1) to say something very different. No longer would this section of the law address the Prodigy situation. Indeed, no longer would it address any situation in which content moderation is the claimed basis for platform liability. By regulatory fiat, the plain meaning of the words in the statute would be erased. Section 230(c)(1) would have “no application” whatsoever to the Prodigy situation that was the occasion for the law in the first place. The Petition contends that erasing the plain meaning of subsection (c)(1) is necessary to give meaning to subsection (c)(2) of the statute. Otherwise, it is argued, (c)(1) is mere “surplusage.”42 As rewritten in the Petition’s proposed FCC rule, therefore, (c)(1) would have “no application” to any case involving content moderation; only subsection (c)(2) could reach such cases. That this does not square with the statute as written is easily demonstrated by imagining that Section 230 did not include subsection (c)(2) at all. What then would be the result in the Prodigy situation? In that case, the language of subsection (c)(1) alone would prevent a court from holding Prodigy liable based on its attempt to enforce content guidelines. The creator of the content, Prodigy’s user, is clearly “another information content provider” distinct from Prodigy itself. Under (c)(1), Prodigy cannot be treated as if it is the publisher or speaker of the content created by that user. Since holding Prodigy liable for its user’s content (whether because of its content moderation policy or for any other reason) would clearly do so, (c)(1) prohibits that result. It is therefore easy to see why courts have consistently found that Section 230(c)(1) can and does, in fact, apply in many cases involving content moderation.43 But now imagine that the Petition’s recommended FCC rule is in place. What then would be the result in the Prodigy situation? Because the rule provides that the entirety of (c)(1) has “no application” to any action involving content moderation, it could no longer be the basis for immunity. This would expose Prodigy to liability on the very ground the New York court advanced in 1995: its adoption of content guidelines designed to restrict the availability of information provided by its users. The proposed rule would dramatically change the meaning of the statute, and the results it produces. But if Section 230(c)(1) can apply to cases involving content moderation, what, then, are the separate, complementary purposes of Section 230(c)(2)? Why is it not “surplusage”?

42 Petition at 29. 43 See, e.g., Barnes v. Yahoo!, Inc., 570 F.3d 1096, 1105 (2009).

12

As drafter of both the (c)(1) and (c)(2) language, Representative Cox wished to ensure that good faith moderation efforts44 were protected even when they might involve an internet platform in what could be deemed content creation. For example, a platform might undertake to revise objectionable content in order to bring it into conformity with its community guidelines, making it responsible for co-creating content, or at least partially responsible for developing the content. The protection in (c)(1) would then be unavailable to the platform, because it extends only to persons that have no involvement in the creation of allegedly illegal content. To address the endless potential fact situations in which content moderation policies could be viewed as participation in content creation or development, Section 230(c)(2) extends protection from liability to any platform that in good faith moderates content it considers to be “obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable.” This protects the platform even if it did create or develop the content, at least in part. It is not “surplusage,” but a very important part of Section 230, and how it is meant to work.45

2. The Meaning of Section 230(c)(2)

The Petition would have the FCC issue rules defining terms in Section 230(c)(2) in ways that impermissibly change the actual language and meaning of the statute. Each of the following revisions of this type that NTIA has proposed is inconsistent with the law as written. a. “Excessively violent”

The proposed FCC regulatory definition of this term would limit it to material that:

is likely to be deemed violent and for mature audiences according the Federal Communications Commission’s V-chip regulatory regime and TV Parental Guidance, promulgated pursuant to Section 551 of the 1996 Telecommunications Act [citation omitted.46

44 Numerous examples of bad faith in content moderation have arisen over the last 24 years since Section 230 was enacted, one of the most notorious being Backpage.com. The operators of that website “moderated” content not for the purpose of removing objectionable material, but rather to disguise it and ward off law enforcement. See Plea Agreement, U.S. v. Ferrer, No. CR-18-464-PHX-DJH (Apr. 5, 2018) at 13, available at https://www.justice.gov/opa/press-release/file/1052531/download (last visited Aug. 30, 2020). 45 Section 230(c)(2) achieves other purposes as well, which the Petition fails to address. See 47 U.S.C. § 230(c)(2)(B). 46 Petition at 37-38. The Petition also includes “terrorism” within its redefinition. For this purpose it borrows the definition for federal criminal law found in 18 U.S.C. § 2331. By so limiting the definition of “terrorism,” the proposed rule would create problems similar to those that would arise from substituting the V-Chip standard of “graphic violence” for the language of Section 230 itself.

13

The statute itself is not so limited. It is unclear why, for policy reasons, NTIA would wish to more narrowly circumscribe the category of violence, effectively preventing content moderation of much horrific content that sadly abounds on the internet. But leaving that question to one side, what is manifest is that the proposed rule would establish a very different, and more narrow, category of objectionably violent content than what is presently in the statute. The statute as written contains its own definition of violent material, which the proposed rule overwrites. As defined in Section 230(c)(2), violent material is that which “the provider or user considers to be … excessively violent.” The Petition would replace this with whatever the FCC considers to be excessively violent, thus changing what Section 230 intends to be many different private sector determinations (each deemed suitable for the particular platform or website) into a single government-issued standard. That single government standard would be lax indeed, waiving in all manner of violent content, which platforms would henceforth be powerless to moderate without fear of liability. The proposed rule’s definition of “excessively violent” would come from the FCC’s V-Chip guidance on what television programs should be rated only for “Mature Audiences.” By limiting Section 230 protection to the moderation of only “graphic” violence that falls in the “Mature Audiences” category, the proposed rule would penalize websites that wish to adopt community guidelines outlawing “intense violence” — because this level of violence is expressly permitted under the V-Chip guidance for the TV-14 category. (“Intense sexual situations” and “strong coarse language” are also acceptable in the TV-14 category.)47 The entire premise of Section 230 is that the wide variety of internet platforms, and internet users themselves, should be empowered to work together to determine what content they find objectionable, and then prevent that content from crossing their portal. A website catering to children, for example, would not wish to host content that is “intensely violent”; similarly, an adult website devoted to discussion of orbital mechanics would likely wish to filter out “intensely violent” content, all of which would be unwanted. The proposed FCC rules would strip such websites of Section 230 protection by eliminating subsection (c)(1) as a source of immunity for content moderation, and then re-defining “excessively violent” in subsection (c)(2) to exclude content that is “intensely violent.” All of this is directly at odds with the statute as written, and it accomplishes no discernable purpose except to prevent voluntary, private activity by websites to keep content they consider excessively violent off of their corner of the internet.

47 See Congressional Research Service, The V-Chip and TV Ratings: Monitoring Children’s Access to TV Programming (2011), Appendix.

14

b. “Harassing”

The Petition seeks a rule restricting the meaning of “harassing” to material that:

has the subjective intent to abuse, threaten, or harass any specific person and is lacking in any serious literary, artistic, political, or scientific value; [is] regulated by the CAN-SPAM Act of 2003, or that is malicious computer code intended (whether or not by the immediate disseminator) to damage or interfere with the operation of a computer.48

As with the proposed redefinition of “excessively violent,” this revision of the statute’s meaning begins by amending it to substitute a single federal government metric for what the statute intends to be many different private-sector determinations tailored to the needs of each website. The proposed rule would effectively strike the words “material that the provider or user considers to be” harassing.49 This is impermissible for a rule purporting merely to “interpret” the law and resolve “ambiguity.” In combination with the proposed new requirement that harassing material be determined on the basis of the content provider’s “subjective intent,” the new definition would make it impossible for a platform to claim Section 230 protection for moderating content on the basis of its judgment of what is harassing. And this, in turn, would mean that courts would be able to rule on such cases at the pleading stage far less frequently — introducing significant new costs for websites that currently benefit from the law’s specification that it is their judgment that matters. The Petition’s proposed substitution of a single federal government standard in place of a flexible private sector standard for what is considered “harassing” suffers from under-inclusiveness. By limiting what can be considered “harassing” to only certain types of spam, and to material “lacking in any serious literary, artistic, political, or scientific value,” it is needlessly restrictive, raising the question of what the NTIA’s purpose could be in making it easier for bad actors to harass people on the internet. That the proposed definition is needlessly restrictive is illustrated by some simple examples. There may well be “serious political value” in harassing elected officials or their families, in which case any platform that tried to control it would lose its Section 230 protection. Organizing a protest to firebomb a person’s home, surrounding a person on the streets with an intimidating flash mob, stalking or threatening their children, all would amount to objectionable harassment

48 Petition at 38 (citations and internal numbering omitted). 49 47 U.S.C. § 230(c)(2)(A) (emphasis added).

15

in any ordinary sense of the word, but each would be left out of the new definition if it had a serious political purpose. Similarly egregious forms of harassment might be placed off limits for protected content moderation because they have artistic merit; there are entire categories of creative works such as “scatological art” that offer innumerable opportunities for online weaponization as harassment.50 Without having to predict that a platform would lose its Section 230 protection in every case because of such definitional exceptions, it is enough to recognize that the proposed rule would create new litigable issues by introducing ambiguities not presently in the statute. Limiting the reach of “harassing” to spam covered by the CAN-SPAM Act is likewise needlessly restrictive. The CAN-SPAM Act does not apply to non-commercial spam.51 If commercial spam can be considered harassing, why not, for example, political fundraising spam? While the government’s restriction of political fundraising solicitations raises First Amendment issues, many people find it attractive indeed to filter such unwanted solicitations, and Section 230 as written clearly allows private platforms to do this. Yet the Petition’s proposed rule would no longer permit websites to filter political fundraising spam and still retain their Section 230 protection from liability for what their users post. Similarly, any other form of spam emanating from any nonprofit organization would fall outside the proposed redefinition of “harassing.” There is no basis in the statute for such a distinction.52

c. “Otherwise objectionable”

The Petition proposes to define the words “otherwise objectionable” to mean:

material that is similar in type to obscene, lewd, lascivious, filthy, excessively violent, or harassing materials.53

According to the Petition, this definition is meant as a restatement of the long-standing canon of statutory interpretation, ejusdem generis, applicable in the case of a statutory list of specific elements followed by general words forming a residual clause.54 In fact, however, the proposed

50 See, e.g., Gabriel P. Weisberg, “Scatological Art,” Art Journal, 52:3 (1993), 18-19. 51 Controlling the Assault of Non-Solicited Pornography and Marketing Act of 2003, Pub.L. 108–187, 117 Stat. 2699, 15 U.S.C. chap. 103. 52 The proposed inclusion of certain types of computer hacking in the definition of “harassing” is unobjectionable in itself, but it needlessly raises the question of whether, by including only this select variety of hacking in the definition, other forms of hacking might lie outside the definition by virtue of exclusio unius est exclusio alterius. See 2A Sutherland Statutes and Statutory Construction, §§ 47.23-25 (7th ed. 2018). 53 Petition at 38. 54 Id. at 32.

16

definition is an improper abridgement of that canon which would significantly restrict the kinds of objectionable content websites can safely moderate without fear of liability. The canon is part of the bedrock of English and American law, dating at least to the 16th century,55 and it is applied in the 21st century in an average of about 90 cases each year in federal and state courts.56 Its operation has been usefully summarized in the leading treatise on statutory construction as containing five elements.

Ejusdem generis applies under the following circumstances: A statute contains an enumeration by specific words; the members of the enumeration constitute a class; the class is not exhausted by the enumeration; a general reference supplements the enumeration, usually following it; and there is no clearly manifested intent that the general term be given a broader meaning than the rule requires.57

Crucially, the Petition’s truncated version of ejusdem generis excises the fifth element. But this qualification, recognizing the importance of a contrary intent or purpose that may be clearly inferred from the context of the statute in which the enumeration and general reference appear, is essential to faithful application of the canon.58 In its 2001 decision in Circuit City Stores Inc., v. Adams, applying the canon of ejusdem generis to interpret a provision of the Federal Arbitration Act, the Supreme Court noted that courts “must, of course, construe [a general term that follows specific terms in a statutory enumeration] with reference to the statutory context in which it is found and in a manner consistent with the [statute’s] purpose.”59 In Circuit City the Court found that this consideration supported the narrower construction of the catchall term that ejusdem generis would ordinarily suggest; but in Norfolk & Western R. Co. v. Train Dispatchers, cited approvingly in Circuit City, the Court relied upon the fifth element of the canon and determined that “the whole context” of the statutory language manifested congressional intent that the general term be given a broad reading.60 And in Arcadia v. Ohio Power Co., involving a statutory provision containing four

55 See Archbishop of Canterbury’s Case, 2 Co. Rep. 46a, 76 E.R. 519 (1596). 56 Preston M. Torbert, “Globalizing Legal Drafting: What the Chinese Can Teach Us About Ejusdem Generis and All That,” 11 The Scribes Journal of Legal Writing 41, 43 (2007). 57 2A Sutherland, supra, §§ 47.18-22. . 58 Id. § 47.22. 59 532 U.S. 105, 118. 60 499 U.S. 117, 129 (1991).

17

specific terms followed by “or any other subject matter,” the Court took pains to examine the surrounding statutory context before concluding that ejusdem generis required a narrow reading of those words (otherwise, the Court said, the phrase “or any other subject matter” would have “swallowed what precedes it”).61 As the NTIA surely is aware, courts applying Section 230 have explicitly treated each of the elements of the ejusdem generis canon, including the qualification that courts must give effect to clear expressions of congressional intent bearing on the interpretation of the general term “otherwise objectionable.” In Enigma Software Group U.S.A v. Malwarebytes, Inc.,62 for example, the Ninth Circuit examined the entirety of Section 230 before determining that a proper application of ejusdem generis “would not support [a] narrow interpretation of ‘otherwise objectionable.’” As part of this examination, the panel noted that the specific categories in Section 230(c)(2)(A) vary greatly, covering a wide array of potentially objectionable material; that “spam, malware and adware could fairly be placed close enough to harassing materials to at least be called ‘otherwise objectionable’ while still being faithful to the principle of ejusdem generis”; and that Congress had taken “the rather unusual step of setting forth policy goals in the immediately preceding paragraph of the statute.”63 Eliminating the fifth step of the ejusdem generis canon, as the Petition’s proposed FCC rule would do, will have the effect of overturning both judicial interpretation of Section 230 and the settled judicial application of the canon itself. Without allowing for the qualification of statutory context and manifest legislative purpose that has long been part of the ejusdem generis analysis, the congressional purpose in adding “otherwise objectionable” will be lost. As the court noted in Enigma Software, Congress stated plainly that its policy objective in Section 230 was to encourage private sector blocking and filtering of “objectionable or inappropriate online material,”64 a deliberately broad category. The Petition seeks to further narrow the scope of content moderation protected by Section 230 by positing a close connection between each of the six specific elements in subsection (c)(2)(A) that precede the words “otherwise objectionable.” The thread that unites the six terms, it asserts, is that each one can be assigned a “regulatory meaning.”65

61 498 U.S. 73, 78 (1990). 62 946 F.3d 1040 (2019). 63 Ultimately the court held that even a broad interpretation of “otherwise objectionable” would not extend to the defendant’s determination that content was objectionable because it would “benefit a competitor.” The catchall term “otherwise objectionable” does not, the court said, give providers “unfettered discretion to declare online content ‘objectionable’.” Id. 64 47 U.S.C. § 230(b)(4). 65 Petition at 37.

18

Thus, the first four terms could all be interpreted with reference not only to the 1873 Comstock Act (a virtual dead letter in the 21st century that was enacted during the Grant administration), but also, by very loose analogy, to the words “patently offensive” included in the now-defunct portion of the Communications Decency Act authored by Senator James Exon and invalidated as unconstitutional by the Supreme Court. This illogical stretch, the Petition all but confesses, is essentially a jurisdictional bootstrap that would amend Section 230 to more closely align it with subject matter that “the FCC continues to regulate to this day.”66 The Petition suggests that the remaining two specific terms in subsection (c)(2)(A), “excessively violent” and “harassing,” could be defined with reference to terms the FCC has used in its V-Chip guidelines. But the concept of “harassment” is entirely absent from those guidelines. As for “violence,” the guidelines mention it but do not attempt to define it. Instead, they perfunctorily categorize it, without any further explanation, as “intense” or, presumably worse, “graphic” violence.67 While this would be of little help to courts and the public in understanding the meaning of Section 230, it would have the (presumably intended) jurisdictional benefit of more closely aligning Section 230 and a subject of regulation with which the FCC has “long been concerned.”68 Finally, the Petition suggests that “harassing” could be defined with reference to federal law that prohibits “mak[ing] or caus[ing] the telephone of another repeatedly or continuously to ring.”69 While this represents a trivially small portion of the vast universe of activity that can constitute harassment on the internet, hunting down repetitive phone callers happens to be an area in which the FCC has existing regulatory authority. Its near total irrelevance to clarifying the meaning of “harassing” in Section 230, however, suggests that the Petition’s actual priority is creating jurisdiction for the FCC rather than statutory exegesis. Unfortunately, such a weak example of the FCC’s “experience” in regulating online harassment serves rather to demonstrate that the agency actually has little. The net effect of this strained effort to compact the six separate, specific categories of subsection (c)(2)(A) into a single concept is, according to the Petition, to demonstrate that “otherwise objectionable” must be limited to “content regulation intended to create safe, family

66 Id. at 34. 67 Congressional Research Service, The V-Chip and TV Ratings: Monitoring Children’s Access to TV Programming (2011), Appendix. 68 Petition at 35. 69 47 U.S.C. § 223(a)(1). See Petition at 34-35.

19

environments.”70 But nothing in Section 230 limits its application to content moderation intended solely to protect families with children. As the Ninth Circuit observed in Enigma Software, the specific categories in (c)(2)(A) in fact vary widely. Fairly read within the context of the entirety of Section 230, as the canon of ejusdem generis requires, the “otherwise objectionable” term should reasonably include such varied additional categories as animal cruelty, encouragement of suicide, undisclosed deep fakes, spam of all kinds, and defamatory material (the latter category having been universally accepted by the courts as included within the scope of Section 230 precisely because of the clearly manifested congressional intent, even though it is not mentioned within the (c)(2)(A) list). And as the Enigma Software court correctly speculated, the drafters of Section 230 also wished to cover future forms of objectionable online content that were not clearly foreseeable in 1996. The rule proposed in the Petition would ignore the manifest intent of Section 230 and strictly limit it in a way that is inconsistent with the proper application of the ejusdem generis canon. In so doing it would jeopardize content moderation efforts aimed at cleaning up some of the worst pathologies infecting the internet.

d. “Good faith”

The Petition seeks to “define” the term “good faith” in Section 230(c)(2)(A) by grafting an entirely new section onto the statute. Rather than providing any sort of definition, this new material would add several specific requirements with which websites must comply. None of these is contained in Section 230. Failure to comply with any one of these entirely new requirements would result in loss of the protection of subsection (c)(2)(A) for the website’s content moderation activities. Under the new regulatory requirements, a platform would have to adopt:

publicly available terms of service or use that state plainly and with particularity the criteria the interactive computer service employs in its content-moderation practices, including by any partially or fully automated processes, and that are in effect on the date such content is first posted ….71

Nowhere in the statute is there even a hint of such a requirement.

70 Petition at 37. 71 Id. at 39.

20

Even if a website were to comply with this new federal regulatory mandate, the proposed rule would still deny it Section 230 protection for its content moderation efforts unless the website also:

has an objectively reasonable belief that the [moderated] material falls within one of the listed categories set forth in 47 U.S.C. § 230(c)(2)(A).72

As written, the statute provides that the determination of whether material is objectionable is to be made by “the provider or user.”73 The rule would effectively amend the statute to remove this language and replace it with “a reasonable person.” And while the actual language of Section 230 also makes clear that the “provider or user” must be acting in good faith in making their determination, the statute plainly states that the determination must be theirs. The test is whether they are honestly motivated in their actions, not whether someone else considers the material to be objectionable or not. The proposed rule impermissibly alters the plain meaning of the statutory text. Even compliance with both of these new requirements, at odds with the text of Section 230, would still be insufficient to qualify a website for the protection of subsection (c)(2)(A). Two further, entirely new, preconditions would be added. First, a website would have to establish that it does not:

apply its terms of service or use to restrict access to or availability of material that is similarly situated to material that the interactive computer service intentionally declines to restrict ….74

Second, the website would have to provide to each of its users whose content has been moderated:

timely notice describing with particularity the … reasonable factual basis for the restriction of access and a meaningful opportunity to respond ….75

Whether content provided by different users is “similarly situated” is, of course, an inherently subjective question, and to answer it could necessitate detailed analysis of many thousands or millions of content moderation decisions, depending on the scale of the particular website. But

72 Id. (emphasis added). 73 47 U.S.C. § 230(c)(2)(A). 74 Petition at 39. 75 Id. at 39-40. There is an exception to this proposed requirement for cases in which the website reasonably believes that the content is related to criminal activity.

21

leaving to one side the desirability of such a test, it is enough to recognize that this requirement appears nowhere in Section 230. Divining it from the words “good faith” strains credulity. Similarly, the imposition of a requirement for due process notice and an opportunity for hearing with respect to every content moderation decision made by a website is an extravagant regulatory addition to the statutory text. It would expose websites to significant expense, necessitating far deeper involvement in researching and analyzing user-created content than is feasible for most of them. Websites will naturally seek to avoid or at least minimize this greater expense. If every content moderation decision triggers a requirement for notice, explanation, and a hearing at which the user responds, then the only way to minimize the associated expense will be to reduce the number of moderation decisions. Since every website will have control over the specifics of its content moderation policy, the incentive will be to minimize the number of moderation decisions required, through the adoption of less robust moderation policies. Alternatively, websites could reduce or eliminate user-created content. Section 230, on the other hand, is intended to protect and encourage content moderation, and to facilitate users’ ability to publish their content on the internet. The inclusion of the notice and hearing requirement lacks foundation in the text of Section 230 and is at odds with the stated goals of the law. That it is masquerading as a “definition” of the term “good faith” only amplifies its illegitimacy.

3. Section 230(c)(1) and 230(f)(3)

The Petition urges the FCC to promulgate a regulation that defines the words “responsible, in whole or in part, for the creation or development of information” in Section 230(f)(3). The proposed definition would include, but not be limited to:

substantively contributing to, modifying, altering, presenting or prioritizing with a reasonably discernible viewpoint, commenting upon, or editorializing about content provided by another information content provider.76

This proposed redefinition of language in Section 230(f)(3) that on its face is plainly understandable introduces new ambiguities rather than resolving any existing ones. Under the proposed redefinition, “contributing” to content must be substantive, but “modifying” and “altering” need not be. This internal logical inconsistency makes for definitional hash; the proposal is at war with itself. It would also overrule the established jurisprudence requiring that

76 Id. at 42.

22

content creation be “substantial”77 or “material”78 (or “substantive,” the Petition’s proposed synonym that modifies “contributing” but not “modifying” or “altering”). Sowing confusion where none currently exists is not an argument for the FCC to promulgate a “clarifying” rule. The balance of the proposed definition introduces still more needless, new interpretive difficulties, rendering uncertain and ambiguous what is now settled. Including the words “presenting or prioritizing with a reasonably discernible viewpoint” is apparently intended to call into question the prioritization of content on a website. But ranking content, whether by page views, by centrality, by some other metric, or even randomly, is a necessary activity for websites in order to make third-party content accessible to users. Without some way to organize the typically large volumes of user-provided content featured on most websites, the sites would become nothing more than shambolic mountains of unusable miscellany. “Presenting or prioritizing content” is inherent in maintaining any website that displays user-created material. In seeking to convert this essential activity into content “creation or development” for purposes of Section 230(f)(3) by conflating it with expressing “an identifiable viewpoint,” the proposed rule veers far from what the statute plainly says. “Presenting or prioritizing” content without materially altering the content itself is not “creation or development” in any ordinary understanding of those words. Were the FCC to adopt a rule to so change the meaning of the words in the statute, it would effectively eliminate subsection (c)(1) protection for millions of websites. This is directly at odds with Section 230 as written. Finally, by adding the term “editorializing about content provided by another” to its proposed definition, the Petition illegitimately seeks to make websites liable for content created by others. Section 230(c)(1) forbids this. While a website is responsible for content that it creates, it is the essence of the statute, expressed in the plain language of subsection (c)(1), that a website cannot be made liable for “content provided by another.”

4. “Treated as a Publisher or Speaker”

Section 230(c)(1) provides that “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” As discussed above, the Petition’s recommended rule would impermissibly amend this portion of the statute to make it inapplicable in any case involving content moderation. In order to further restrict the application of the law’s plain language, the Petition also proposes an extensive revision of what it means to be “treated as the publisher or speaker of any information provided by another information content provider.”

77 See, e.g, Batzel v. Smith, 333 F.3d 1018, 1031 (9th Cir. 2003). 78 See, e.g., Fair Housing Council of San Fernando Valley v. Roommates.com, 521 F.3d 1157, 1168 (9th Cir. 2008)(en banc).

23

Under the terms of this proposed revision to the statute, a website would not be deemed to be treated as the publisher or speaker of any information provided by another (that is, it would not be protected from liability) whenever:

it affirmatively solicits or selects to display information or content either manually by the interactive computer service’s personnel or through use of an algorithm or any similar tool pursuant to a reasonably discernible viewpoint or message, without having been prompted to, asked to, or searched for by the user ….79

Restating subsection (c)(1) to strip liability protection for selecting which user-created content to display “pursuant to a reasonably discernable viewpoint” effects a change to this portion of the statute that parallels the proposed redefinition of the words “responsible, in whole or in part, for the creation or development of information” in Section 230(f)(3), discussed above. It suffers from exactly the same defects and is, to the same extent, inconsistent with the statute as written. Finally, the Petition recommends a further change to Section 230(c)(1) that would deny liability protection to a website if it:

reviews third-party content already displayed on the Internet and affirmatively vouches for, editorializes, recommends, or promotes such content to other Internet users on the basis of the content’s substance or messages.80

The proposed rule would deny liability protection whether the promotion of the content was done “manually … or through use of an algorithm.”81 Thus, for example, an automated content prioritization tool that displays customized content to each user based on their previous viewing history would run afoul of the rule. There is no indication in the text of Section 230 that this should be a basis for a website to lose its protection from liability for content created wholly by others. These proposed regulatory re-interpretations of the meaning of “treated as a publisher or speaker” are at variance with the actual language of Section 230. As do the other proposed regulatory revisions of Section 230 discussed above, they violate the plain meaning of the statutory text.

79 Petition at 46. 80 Id. 81 Id.

24

D. The FCC Should Exercise Its Discretion to Reject the Petition for Rulemaking

In the event that, despite the clear expressions of congressional intent to the contrary and the statutory text stating Section 230’s purpose of keeping the internet free from federal regulation, the FCC believes it may have the authority to engage in a rulemaking on Section 230, the FCC should nonetheless exercise its discretion to dismiss the Petition for prudential reasons. There are several sound public policy reasons for doing so, each of which advances the FCC’s mission. First, taking on the role envisioned by the Petition will undermine the ability of the FCC to operate as an independent agency. On the basis of an express directive from the president, from which the Petition originates,82 the Commission would place itself in the middle of the most contentious social issues of the day. Worse, doing so would require the FCC to act as arbiter of speech codes on the internet. The Petition specifically calls for the FCC to become involved in deciding when “online platforms ‘flagging’ content” is acceptable and when it is not.83 It urges the FCC to referee whether platform moderation decisions “have the effect of disfavoring certain viewpoints.”84 And it would thrust the Commission headlong into unwinnable arguments over whether platforms can “fact-check” candidates and elected officials.85 In addition to compromising the FCC’s political independence, a second reason for dismissing the Petition on prudential grounds is that were the Commission to assume the mantle of internet speech arbiter that the Petition has envisioned, this would further weaken the FCC’s position in the courts. As the Supreme Court has made clear, the FCC has never been entitled to “a free hand

82 Exec. Order No. 13925, 85 C.F.R. 340179 (2020). Because the Executive Order expresses the president’s concern that some platforms are “disfavoring certain viewpoints” based on political bias, placing the FCC in the role of arbiter of such questions would necessarily require it to resolve inherently political questions. As administrations change from one party to another, this will generate accusations of political bias on the part of the FCC itself. Such suspicions have already risen anticipatorily: former Senator Rick Santorum, writing concerning the Petition, has worried publicly that were the NTIA’s recommendations to be adopted, the result “would not go well for the President or for conservatives.” Since in a future administration the very people who now “demand that social media sites crack down on posts by the President and conservatives” would be in a position to influence the FCC, the Commission’s newfound regulatory power could be used for just such purposes. Rick Santorum, “President Trump Should Bend — But Not Break — Big Tech,” The American Spectator (Aug. 5, 2020), available at https://spectator.us/trump-bend-break-big-tech/ (last visited Sept.1, 2020). 83 Id. at 7. 84 Id. 85 Id. at 8.

25

to vindicate its own idiosyncratic conception of the public interest or of the requirements of free speech.”86

The agency’s authority to regulate speech in the broadcast context, its only direct experience with nuanced content regulation thus far, was upheld on the ground that broadcast spectrum is scarce.87 No such argument exists for regulating speech on the internet, where access to being a “broadcaster” is open to every individual internet user — as is the ability to create one’s own “interactive computer service.”

Even in the area of broadcast media, the Supreme Court has cautioned that the Commission operates on a tight leash when it comes to speech regulation. Any number of actions the FCC might take — for example, refusing to permit a broadcaster to carry a particular program, or to publish certain views; discriminating in determining which broadcasters will be required to broadcast certain views; censoring a particular program contrary to the statute; or advancing an official government point of view —would, in the Court’s view, raise First Amendment issues.88 Beyond this, the Commission’s own experience with the fundamentally flawed Fairness Doctrine89 should reinforce the constitutional concerns with pragmatic ones.

The Petition is uninterested in any of these concerns. Couched in the language of ensuring “freedom of expression,” what the Petition actually is asking the FCC to do is mandate speech. In urging the FCC to promulgate rules mandating that platforms publish content that otherwise they would moderate according to their published terms of service and community standards, the Petition would set the Commission on a collision course with governing First Amendment norms, most particularly the principle that freedom of speech prevents the government from dictating mandatory speech.90 As Chief Justice Burger stated in Wooley v. Maynard, “The right to speak and the right to refrain from speaking are complementary.”91

The third reason that the Commission should exercise its discretion to dismiss the Petition is that upending settled judicial interpretations of Section 230 will unfairly prejudice reliance interests. The millions of websites governed by Section 230 represent a wide variety of e-commerce models that have been built on the basis of the law’s protections against liability for third party

86 Red Lion Broadcasting Co. v. FCC, 395 U. S. 367, 395 (1969). 87 Id. at 394. Thirty years later, in United States v. Playboy Entertainment Group, 529 U.S. 803 (2000), the Supreme Court ruled that this rationale did not apply to cable television, since scarcity was not an issue in that market. 88 Red Lion, 395 U.S. at 396. 89 In the Matter of Editorializing by Broadcast Licensees, Docket 8516, 13 FCC 1246, App. (June 8, 1949), available at https://docs.fcc.gov/public/attachments/DOC-295673A1.pdf (last visited Sept. 1, 2020). 90 See Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston, 515 U.S. 557 (1995); Wooley v. Maynard, 430 U.S. 705 (1977). 91 Wooley, 430 U.S. at 714.

26

content. Invalidating the settled understanding of how Section 230 operates will have significant negative effects on a wide variety of web platforms, and in particular, small businesses. The Supreme Court has specifically cited the importance of reliance interests in explaining the circumstances in which it will invalidate agency rules that break with settled statutory interpretations.92 Fourth, the FCC has regularly avoided wading into the thicket of free speech debates, such as when it decided not to take action against Fox News based on the dozens of complaints accusing the station of “indecency,” “misrepresentation,” and “video editing.”93 Likewise, the FCC chose not to take action against Stephen Colbert and CBS for using obscenity to describe President Trump on the CBS Late Show.94 Exercising similar discretion in this case will be in keeping with the agency’s traditional avoidance of regulatory actions that could be found to restrict expression protected by the First Amendment. Fifth, engaging in rulemakings on the content moderation practices of platforms under Section 230 will open up avenues for abuse of the process by competing political interests for political purposes. It will of necessity place the moderation of online content in the hands of political appointees at the Commission, so that notwithstanding the FCC’s painstaking efforts to remain independent and to avoid the appearance of politicization of its actions, those efforts will be compromised. Political oversight of online speech risks the use of FCC authority both to mandate favored speech and to suppress disfavored speech.

Sixth and finally, adopting the re-interpretations of Section 230 urged in the Petition will be costly to the economy, as well as to millions of unemployed Americans, at a time when the nation can least afford it. As reported in a 2017 study by NERA Economic Consulting, internet platforms and e-commerce sites governed by Section 230 “are a driving force of the modern U.S. economy. Numerous studies document the importance of these and many other entities both domestically as well as internationally.”95 For several years, the internet sector has been growing

92 Encino Motorcars, 138 S. Ct. at 2125-26. 93 Federal Communications Commission, Response to FOIA Request, No. 2017-587 (June 6, 2017) (documents detailing complaints to FCC requesting action against Fox News Channel), available at https://static1.squarespace.com/static/5722daf11d07c02f9c1739cc/t/5938e303e3df28736bd908aa/1496900365579/FCC-complaints-foxnews.pdf (last visited Aug. 29, 2020). 94 “FCC will not take action against Stephen Colbert over controversial Trump joke,” Time.com (May 23, 2017), https://time.com/4791572/fcc-stephen-colbert-late-show-donald-trump/ (last visited Aug. 28, 2020). 95 Christian M. Dippon, Ph.D., “Economic Value of Internet Intermediaries and the Role of Liability Protections,” NERA Economic Consulting (June 5, 2017), available at https://internetassociation.org/wp-content/uploads/2017/06/Economic-Value-of-Internet-Intermediaries-the-Role-of-Liability-Protections.pdf (last visited Aug. 29, 2020).

27

at double digit annual rates, including throughout the Covid recession, proving to be the most significant positive for the nation’s economy during these difficult times.

NERA Economic Consulting estimates that weakening Section 230 could cost the U.S. economy approximately 4.25 million jobs and $440 billion in GDP over a period of 10 years.96 They find that Section 230 is responsible for two to three times more investment in the United States than in comparable sectors in the EU.97 While these statistics demonstrate that the aims Congress expressed in enacting Section 230 continue to be achieved — it is undeniably the case that the “Internet and other interactive computer services have flourished, to the benefit of all Americans, with a minimum of government regulation”98 — it is equally clear that undertaking a regulatory rewrite of Section 230 risks undermining this enormous progress. Pursuing the course recommended in the Petition would also upend one of the original congressional purposes in enacting Section 230: “to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation”99 — including in particular regulation by an FCC seeking to become a Federal Computer Commission.

IV. Conclusion

For the foregoing reasons, the Commission should dismiss the petition.

Sincerely,

Christopher Cox Carl Szabo Christopher Marchese Director Vice President and Policy Counsel NetChoice General Counsel NetChoice NetChoice

96 Id. 97 Id. 98 47 U.S.C. § 230(a)(4). 99 Id.